Hello Chaps,

I am getting resistance values all over the place for R4 and R5 resistor on my PCB. funny enough when I cut the trace from R5 going to junction of c2 where its getting 3.3V, my resistor values become stable and correct. I am wondering is there some sort of ground loop on my pcb ?


For refrence the value of C5 is 470 uF.

 

You cannot measure resistances while in-circuit and expect to get reliable readings.

 

You cannot measure resistances while in-circuit and expect to get reliable readings.

I do that when power is switched off. I know, I wont do it whilst power is on

 

Resistance is calculated by injecting a small voltage (usually 0.5-1.5V) into the circuit and measuring the current. There are other methods, but this
is the main one. IF there are other components, anywhere in the circuit (ie parallel path connection) - ESPECIALLY capacitors, all you are doing
is charging them up (and discharging them when you reverse the leads). Also, IF the test voltage is above 0.6V, you may even be turning on PN
junctions. In your case, with 1M / 470K resistors, the issue is exasperated ie the current is very low and easily swamped.
Measuring resistors in circuit is very tricky.

 

I do that when power is switched off. I know, I wont do it whilst power is on

Removing power is not enough.
You have to disconnect at least one leg of the resistor from the circuit.

 

Removing power is not enough.
You have to disconnect at least one leg of the resistor from the circuit.

Ok, but all the other resistors on the board are showing correct values. Only these two having the issue ?

 

show the schematic and not a partial board layout.

 

Ok, but all the other resistors on the board are showing correct values. Only these two having the issue ?

Are you paying attention?

A resistor in-circuit has the entire rest of the circuit as a baggage and burden on that single resistor.
What you read on a resistance meter will depend on the individual resistor and the size of the baggage and burden hanging on to that resistor.

Do you get it?

 

Resistance is calculated by injecting a small voltage (usually 0.5-1.5V) into the circuit and measuring the current.

Other way around with modern meters. Resistance is measured by applying a current and measuring the voltage.

 

show the schematic and not a partial board layout.

 

Remove C5 from the circuit and repeat the resistance measurements.

 

Remove C5 from the circuit and repeat the resistance measurements.

I already did, removed C5 but it didnt make any difference.

 

I already did, removed C5 but it didnt make any difference.

Did you disconnect the board from everything else that connects to it?

What else is connected to VSENSE?

 

Did you disconnect the board from everything else that connects to it?

Yes

 

Yes

Edited post:

What else is connected to VSENSE?

 

Edited post:

What else is connected to VSENSE?

Its been fed to a MCU.

 

Its been fed to a MCU.

Then you have to break the connection to the MCU.

 

I always take at least one leg off to measure a resistor. That way I feel sure of the measurement.

 

If you attempt to measure the resistance of R5 with the MCU connected, you are literally powering up the MCU through its input pin.

Even with the input pin disconnected from VSENSE, you are still powering the MCU via the supply and ground rails.

You need to remove the MCU from the circuit board.

 

If you attempt to measure the resistance of R5 with the MCU connected, you are literally powering up the MCU through its input pin.

Even with the input pin disconnected from VSENSE, you are still powering the MCU via the supply and ground rails.

You need to remove the MCU from the circuit board.

Got it, it makes sense.